The Appeal of Minimum Income Programmes in Latin America1

Lena Lavinas2

Abstract: This paper aims at describing some minimum income schemes implemented in Latin America in the 1990s. Special attention will be given to school grant programmes – usually called Bolsa-Escola – adopted in Brazil. We also provide information on the ongoing debate on minimum income programs taking place in Argentina and Ecuador and offer a brief description on Progresa, the Mexican version of school stipends. Despite their compensatory bias, minimum income programmes play an important role in expanding the citizenship of the less favored that have been left out from the existing social protection system. More than a safety net, these cash transfers actually have a redistributive impact that is non-existent in the traditional compensatory programmes, since they bring into the debate on the restructuring of welfare the idea of basic security for all. In doing so they provide the necessary framework for moving from selectivity to universalism.

1 Este artigo baseia-se num conjunto de pesquisas sobre programas de garantia de renda mínima

no Brasil, realizadas no IPEA. Ver a esse respeito, Lavinas L. e Varsano R., Programas de Garantia de Renda Mínima e Ação de Combate à Pobreza, Rio de Janeiro: IPEA (Discussion Paper. 534), 1997 and also in Lobato A. L., Garantia de Renda Mínima – Ensaios e Propostas, Brasília: IPEA, 1998); Lavinas L. e Varsano R., Renda Mínima: Integrar e Universalizar, in Novos Estudos CEBRAP, São Paulo, n. 47/1997 p.213-229; Lavinas L. Programas de Garantia de Renda Mínima: experiências brasileiras, Rio de Janeiro: IPEA (Discussion Paper 596) 1998; Lavinas L. Renda Mínima: práticas e viabilidade, in Novos Estudos Cebrap, n. 53, março de 1999, p. 65-83; Lavinas L., Rocha S., Varsano R. Programas de Garantia de Renda Mínima: uma orientação aos municípios. Rio de Janeiro: Ford Foundation/ FINEp/Friedrich Ebert Foundation, 1998, 88 páginas. Contou com a valiosa colaboração das economistas e assistentes de pesquisa Danièle Manão e Mariana Bittar, a quem agradeço.

2 Lena Lavinas is Senior Economist in the InFocus Programme on Socio-Economic Security, at the ILO.

1

1. The legacy of the social protection system in Latin America

The social protection system in Latin America is characterized by its failure in achieving full, universal coverage. Although inspired by the Welfare State model, the Latin American systems have many distortions and deficiencies, which Draibe3 attributes to the characteristics of the structure of the social protection system itself, to the low progressivity of social expenditure, as well as to other causes, like the structure of the labour market and level of wages.

Due to the high degree of informality in the labour market prevalent in many

countries of Latin America, most workers have no access to the social protection mechanisms associated with belonging to the formal labour market, like unemployment benefits or pensions. On the other hand, real wages have always been kept low by state intervention in controlling the trade unions and regulating wages in the private and public sectors. Restrictive regulations towards organized labour have reduced the bargaining power of workers and hindered the appropriation by labor of the productivity increases that occurred in periods of high rates of economic growth. Therefore, the average level of wages remains very low, notably in the less organized sectors.

An even more damaging feature plaguing the social protection system is the

entrenchment of clientelism in Latin American political relations. As a result, certain socio-economic groups have been privileged, at the expense of the majority, mostly poor and deprived of basic rights. Lo Vuolo4 et al. claim that inclusion in the system was directly related to a combination of factors such as type of occupation, level of income and association (“connections”) with specific pressure groups. Unfortunately this corporatist system prevailed up to very recently, when its iniquity and inefficiency in reversing the high levels of social inequality began to be questioned.

Besides the structural features of the Latin American social protection system, that

offers low coverage, the continent witnessed during the 1990s a strong deterioration in social services such as health and education, which was due both to the fiscal adjustment necessary to stabilize the inflation process and to the ineffectiveness of the provision of these services. As more importance was placed in the adjustment, the universal dimension5 of public policies suffered, leading to an increase in poverty and social exclusion. The design of social programmes became subordinated to their financing structure, and principles like justice and equity were abandoned in favor of efficiency and selectivity.

3 Draibe, S. (1997). “A Política Social na América Latina: o que ensinam as experiências recentes

de reformas?” In Reformas Setoriais de Impacto Estratégico, p. 213-257.

4 Lo Vuolo, Rubén (coord). Pobreza y Politicas de Sostenimiento del Ingresso. Forthcoming in 2000, Buenos Aires: CIEPP, 463 p.

5 Nas palavras de Draibe, S. , op. cit, “a perda de participação relativa do gasto com educação e saúde significa o enfraquecimento dos programas universais, de maior potencial de cobertura e que, afinal, expressavam o maior esforço de desenvolvimento social e extensão da cidadania realizado através da ação estatal (p. 249).

2

The social policy model for poverty alleviation programmes has been denominated by Lo Vuolo6 as being “target assistencialism” as opposed to one of universal principles. The current debate is between those that defend the targetting of social expenditure to ensure that the more vulnerable and less represented groups can guarantee their citizenship, and those that believe that it can only truly exist if it broadens the scope of the social protection system. Nevertheless both sides agree that it is definitely urgent to undertake an in-depth reform to redesign the social protection system as a whole, since its ineffectiveness has far reaching consequences on increasing poverty and social exclusion.

The following figures illustrate the relevance of these issues. In Brazil the official

estimates of the number of poor are between 35 and 55 million,7 depending on the poverty line adopted. In Argentina this number corresponds to 20% of the population,8 amounting to about 8 million poor. In Chile, in spite of the improvement in the poverty indexes in the late 1990s, the living conditions of 23% of the population9 are far from decent, and 6% are indigents, making up a grand total of 4.1 million people. In Mexico there are about 5 million families living in extreme poverty. In summary, it has been estimated that, in broad terms, about half of the Latin American population can be considered poor.10

In the midst of this debate several original initiatives to reform the social protection

system, in particular its compensatory side, have surfaced, some of which are inspired on the concept of guaranteed income. They are minimum income programmes, which mostly take the form of a school grant, providing cash transfers to poor families with children of school-going age. They are particularly attractive because in addition to having a positive impact in reducing poverty they increase educational attainment and contribute to the elimination of child labour.

The experiences with minimum income programmes (MIGP) began in the developed countries this century as the Welfare State gained ground. The idea, already formulated by liberal thinkers since the eighteenth century, was to create a network of social protection for the poor by means of a complementary income transfer. Many European countries by the 1930-40s had begun to adopt such programmes. Nowadays, the European Union and OECD recommend the adhesion of the member countries to minimum income programmes, even though such a decision is in the sphere of national sovereignty. Hence the wide diversity of MIGP formats existing today. They all, however, have several points in common11: (1) they are directed to all that are in need, due to insufficiency of income, (2) they are a subjective right, assigned on the basis of a

6 Lo Vuolo, Rubén (coord). Pobreza y Politicas de Sostenimiento del Ingresso, op. cit.

7 IPEA Comissão Especial do Senado Federal pela Erradicação da Pobreza, December 1999. Relatório Final da Comissão

Mista Especial destinada a Estudar as Causas Estruturais e Conjuntuais das Desigualdades Sociais, Comgresso Nacional, Brasília,

1999, 129 p.

8 Censo Nacional de Población Viviendas, INDEC, 1991.

9 Fonte MIDEPLAN, División Social, Departamento de Estudios Sociales, com base nas Encuestas CASEN 1987 -1996.

10 Bruno Sander, Institucionalidad Social para la Superación de la Pobreza y Promoción de la Equidad”. Seminario Taller,

OEA/MIDEPLAN/CEPAL, Roland Franco Editor, 1998, p. 15.

11 Lavinas L. e Varsano R. Programas de Garantia de Renda Mínima e Ação Coordenada de Combate à Pobreza, op.cit.

3

demand made by the individual himself, (3) very often, they are a conditional right, subject to some requirements, and (4) they are a subsidiary right, since the value of the benefit depends on income and access to other social protection mechanisms.12

The minimum income is a cash transfer from the State to individuals or families

that are deprived of minimum living conditions. Generally speaking, the amount of minimum income is the differential between the sum of all incomes of a family and the minimum sum required to survive, this ceiling stipulated on the demographic composition of the family, that is, number of children, inactive or elderly members, and the living conditions in the country. The income value to be transferred is calculated on an accumulative basis. Thus, the sum paid to an adult alone is the reference value based on which the fraction to be paid to the other family members is calculated. This figure is, therefore, below that which a family of two adults without children should receive, which in its turn is also below that to be paid to a family consisting of two adults and one dependent, and so on. There is no pre-fixed sum per family, the sum transferred is calculated on a case by case basis.

With regard to the conditions of admission to the system, they are almost always restricted to the income deficit, implying transparency in declaring the value of the actual income and assets of each candidate. Pre-conditions exist, however, ranging from unconditional to almost obligatory, vis-à-vis active unemployed adults, in accepting any work offered by the public employment system (workfare). These programmes are all restricted to the needy population. It is true that the idea of granting a universal citizen income is under consideration in certain countries like Finland, Holland, Ireland and Belgium. In most countries, however, the predominant scheme is still a compensation mechanism justified by an acute monetary deficit, which adds to other universal benefits allocated to the most impoverished citizens.

Meanwhile, in certain Latin American countries the compensation policies in the

form of guaranteed minimum incomes are very recent. They are not part of the array of assistance programmes whose length and range have always been liable to the multiple forms of political paternalism. It is also true that in most countries of the South American continent universal allocations like family or housing allocations do not exist, and when they do, their monetary value is absolutely derisory not to say immoral. The most widespread compensatory programmes are invariably based on in-kind distribution, with no effect on poverty reduction. That’s why compensatory programmes are acknowledged to be less than effective and/or efficient, corroborating evidence that this compensatory approach adds little to the welfare of social groups suffering extreme poverty. Such ineffectiveness ends up discrediting this type of intervention and is increasingly viewed as needless by society at large. However, the scope of the destitution still jeopardizing Latin American development today forbids one to rule out compensatory measures that might truly reduce poverty and guarantee a basic threshold for citizenship. On the contrary, it is crucial to provide such measures which will form the basis of a new framework for them to truly achieve their objectives and goals by gaining effectiveness.

12 Euzéby, Chantal. Le Revenu Minimum Garanti, Paris: La Découverte, collection Répères, 1991., cap. III.

4

Guaranteed minimum income programmes appear to provide such a new framework. They have emerged as a more effective alternative for reducing poverty in the short run by providing stipends to poor families, and in the long run by increasing educational achievement, since they take into account the strong inverse correlation between educational level and poverty rates. It is important to bear in mind that schooling gap rates remain extremely high in these countries and that they are very significant in explaining poverty levels. To illustrate this fact, note the current average educational differentials between various Latin American countries, focusing on the lower deciles in income distribution (Graph 1). Brazil has the worst performance, with its adult population having an average of 5-6 years of schooling, falling short of completing primary school. In addition, among the poorer population strata the problem of deficient schooling is even worse, underscoring the unequal access to universal lay education in countries like Brazil, Mexico, Paraguay, and Ecuador.

Graph 1

ADULT AVERAGE YEARS OF SCHOOLING(OVER 25 YEARS OLD)

0

2

4

6

8

10

12

14Decile 1

Decile 2

Decile 3Decile 10

Total

Equador

México

Argentina

Brasil

Paraguai

Uruguai

Chile

Source: BID. Facing up to Inequality in Latin America, 1998-99.Prepared by Lavinas, L. and staff, IPEA 2000.

These programmes also help to combat child labour, a distortion that reaches alarming proportions in some countries of Latin America. For example, in Brazil there are 2.9 million children of age 10-14 years and 4.7 million children of age 15-17 years that work.13 These represent activity rates of 16.9% and 45.9% respectively, which means that one in three working children do not attend school. In Mexico, the figures are even higher.

13 Sabóia J. and Sabóia A. L. Trabalho de Crianças e Adolescentes no Brasil dos Anos 90, miméo, 1999, 46 páginas. Report to the World Bank and UNICEF.

5

In order to have an idea about the existing schemes of minimum income programs

lately undertaken in Latin America, we will describe in detail the Brazilian experience and collate some information on similar experiments launched in Mexico or under discussion either in Argentina and Ecuador.

2. Decentralized experiments in Brazil

The second half of the 1990s in Brazil brought various important changes in social policy. With the strengthening of the democratization process in the country, particularly through state, municipal and federal government elections, and with the decentralization process adopted in 1998 by the new Citizens’ Constitution providing greater fiscal and political-administrative autonomy for the sub-national levels, the debate concerning the social inequalities of the country gained a new lease of life and also a new institutional framework.

At the same time, disinvestment by the federal government, previously

responsible for a major portion of the funding and implementation of social policies, required an almost immediate response by local and state governments to deal with the degradation of all types of social services. Therefore, per capita social expenditures increased at the local level, especially in the most developed and strongest metropolitan areas (in the South), that is those with a relatively large and diversified fiscal base. The assessment that besides income insufficiency, poverty is also derived from the acute deficit in the provision of essential public services like health, education, housing and basic sanitation raised serious doubts about the traditional ways of formulating social policy. The popular perception that the country spends a reasonable amount in the social area (19% of GDP), but spends it badly and that public resources never reach the most deprived groups gave way to the adoption of innovative experiments in the more progressive local administrations.

The participatory budget14 thus began to appear in a number of left-leaning

municipal governments,15 where through a concerted effort between civil society organizations and government it was sought to increase public involvement over the allocation of a growing part of public resources. Food safety programmes centered both on the subsidized sale of food to high risk groups and on the regulatory intervention of prices have multiplied; programmes to guarantee a minimum income to the impoverished population have been adopted as a way of alleviating poverty and reducing inequality.

The idea of providing a minimum income for the poor segment of the population was introduced16 in Brazil in the 1970s, but only in the 1990s did it really become an important issue in the national debate on combating inequalities, when Senator Eduardo Suplicy, of the Worker’s Party, presented a bill of law to introduce a guaranteed

14 In the early 1990s progressive municipalities began to set aside a part of the budget for public debate and proposals.

15 Like Porto Alegre (capital of the State of Rio Grande do Sul), Santo André (in the State of São Paulo) Belo Horitzonte (capital of Minas Gerais).

16 See in this respect the articles of Prof. Antônio Maria da Silveira.

6

minimum income programme for all Brazilian adults (over 25), with a monthly income below R$ 240 (US$141). His proposal was to grant a monthly benefit equal to 30% of the difference between the income of the needy individual and the above value, instituting a type of national poverty line. Although the bill was passed by the House of Representatives, it never came to a vote in the Senate, where the debate was blocked by the government majority. Nevertheless, there have been numerous favorable reports issued by the various Congressional committees charged with reviewing the issue.

Set against this proposal of a minimum income is another form of monetary income transfer directed not at poor individuals but at poor families with children of school-going age. This initiative, the work of the economist José Marcio Camargo, proposes granting a monthly benefit of one minimum wage to every family, irrespective of income, whose children are enrolled in public elementary schools. The argument in favour of this form of monetary income transfer is based on the understanding that one of the major underlying factors behind the continuing reproduction of vicious circles of poverty in Brazil is the lack of schooling. In fact the school gap still remains high in Brazil, although declining. According to the 1998 PNAD,17 the average 14 year-old, who should have completed schooling (8 years) has a 2.9 year lag. The adult Brazilian population (over 25) has an average 5.6 years of schooling and only the capital of the country, Brasília, recorded in 1998 a satisfactory performance in terms of completing basic schooling: on average, its inhabitants completed the eight years of compulsory schooling which takes one through primary and middle school. Obviously, looking at the poorest segment of the population, the indicators are even more dramatic, lying well below18 the national average and revealing serious accumulated disadvantages. As an example, one could cite the case of Recife, a metropolitan area in Northeastern Brazil, where children aged 14, living in families whose per capita income is below half a minimum wage (R$60, US$37), recorded a 4.3 year school lag in 1997, while for those from families with an income of less than 1/3 of a minimum wage (US$ 24), the lag was 5 years.19

The first Brazilian minimum income programme appeared in Greater Metropolitan Campinas, in the most highly developed state of Brazil, São Paulo, in 1995, as an additional component of the local social assistance network with 2,500 recipient families by the end of 1999. The format of a guaranteed minimum income tied to compulsory school attendance, was implemented successfully in Governor Critóvam Buarque’s20 administration in the Federal District, in 1994. At the end of his mandate in 1998, the Bolsa-Escola Programme had reached a total of 26,000 families, nearly 80%21 of the potential target population, selected on the basis of a poverty line of half a

17 National Survey by Household Sampling carried out annually by the Brazilian Institute of Geographic and Statistics in 20,000 households (national sample).

18 It should be mentioned that the first guaranteed minimum income programme appeared in the city of Campinas, in the interior of São Paulo State, in 1994.

19 See in this respect I Preliminary Report of the Research “Assessment of Two Bolsa-Escola Programmes,” March 2000, IPEA/OET/BM, coordinated by Lena Lavinas, p47.

20 Elected in 1994 for the Workers’ Party, as governor of Brasília. 21 This high degree of coverage is due to the fact that the eligibility criteria in the Programme have excluded

families resident in the Federal District for less than 5 years, so as to avoid “importing poverty” from neighboring municipalities.

7

minimum wage per capita (US$38). For the first time, a social programme reached a sufficiently large scale and coverage to generate an effective impact on the impoverished population. The monthly benefit of one minimum wage R$120 (US$76) in a direct transfer of monetary income allowed more than 10,000 families to escape from abject poverty and helped focus social expenditure on actions for combating poverty.22 Per capita monthly income between 1995-97 went from R$113 (US$78) to R$279 (US$168) while general social expenditure remained at around R$450 (US$281). The school drop-out rate fell to zero among the recipient families and the repeat student rate dropped well below the average for the city. All this, with less than 1% of the annual budgetary resources of the Federal District.

In view of the very satisfactory results and low operational cost, and in the absence of perverse trade-offs that usually have an adverse effect on the efficiency of the majority of social programmes, the Federal District bolsa-escola programme has become something of a model in Brazil.

Today, a good hundred municipalities23 have adopted this decentralized model, unfortunately, however, without heeding certain important aspects of its design and the way it is implemented, executed and monitored. According to a survey by IPEA24, as of January 2000, there were 57 programmes in operation, almost all of which are run by local governments (45 of them are in the interior of the state of São Paulo). Most provide scholarships25 with stipulations for participating families, such as regular school attendance by children age 7-14 years. A total of about 100,000 families are reached by these programmes.

In most cases, these programmes satisfy only a very small part of the demand,

generating inequalities among the poor population itself. They provide an almost derisory sum for the benefit – when they do not replace the monetary benefit with food, kitchen gas supply or other types of inkind donation – with virtually a zero impact in reducing short-term poverty. They also do not guarantee continuity, as they are committed to maintaining the benefit for only one or two years, since there are no clearly defined targets to be reached or coordinated from the standpoint of social policy as a whole. This yields a negative effect on the utility of the instrument. The impact is therefore lower and it will be even more difficult to overcome the problems of poverty among the needy in the long run. Almost all the municipal or state laws for adopting bolsa-escola programmes do not make the completion of basic schooling their goal, but are only concerned about “removing children from the streets”, which reinforces the social assistance character of these programmes to the detriment of their redistributive effect.

22 See in this respect Lavinas, L. “O Barato da Pobreza”, in Inteligência, year 1 no. 4,

August/September/October/1998, p36-46 . 23 Brazil has about 5,800 municipalities. 24 See Lavinas L. and Bittar M. “Obtaining Educational Improvement Through Basic Income Policy in

Brazil”, mimeo, Rio de Janeiro: IPEA, 2000. 25 There is a so-called “family wage” in Brazil, a tiny stipend paid by employers, i.e., a minor income transfer

to workers in the formal sector with preschool or school-age children (stipulating proof of their children’s school enrollment). In practice, it has never acted as a financial incentive to keep children in school, nor has it served as a scholarship. The monthly value of this stipend is based on the wage level, varying from R$8.00 for lower wage brackets to a few cents for workers earning more than 20 times the minimum wage.

8

There are cities which practise a rotation of recipients in view of budgeting restrictions: some families are granted the benefit for a year, making way at the end of this period to other families. Thus some start to win while others start to lose again, in a zero sum game, where everything is temporary, except the misery.

On the eve of the 1998 presidential elections, the Ministry of Education also launched a guaranteed minimum income programme, geographically based26: municipalities with per capita income and tax revenues below the average for their state could form a partnership with the federal government, which would fund half the costs of a bolsa-escola programme, once again linking the receipt of the benefit to compulsory school attendance. Such exclusionary criteria eliminate some 4.1 million poor families27 who reside outside the target areas, penalized for living in wealthier cities28 and/or for not having school-age children. The distortions stemming from this territorial focus are worrying, since municipalities with a limited tax base often display high poverty rates, given their low level of economic development. To cover the eligible poor families would thus commit an extremely high (and thus prohibitive) share of the local revenue, generating trade-offs between the degree of coverage and the amount stipulated for the benefit. There is thus a major risk of such a program having a zero impact. We have performed simulations29 showing that three-fourths of all Brazilian municipalities would have to commit from 5% to 50% of their current revenue to ensure benefits of 15 reals (less than US$10) a month to all families with incomes below half the minimum wage, with at least one dependent child in the 0-14 age group. The situation is obviously more dramatic in poorer regions like the North and Northeast, where 96% of all municipalities would have to commit from 5% to 50% of their current revenue to reach the targeted families. In other words, the option is simply not feasible.

In December 1999, according to Ministry sources, 504,000 families in one-fifth of

all Brazilian municipalities (located in rural areas in the main) received a monthly benefit whose value varied, but it is estimated on average to be R$37 (US$20). However, estimates as to the number of families in the guaranteed minimum income target group based on established criteria indicate a potential demand of 7.2 million families30, underscoring the limited coverage (only 7%) achieved thus far by the federal programme. Nonwithstanding, more than a million children are thought to have participated in the programme. It is not known, however, how far this programme is working towards

26 See appendix 1. 27 Estimates by Lavinas L., 1998 – IPEA/DIPES – see Appendix 2. 28 Some 3,400 of Brazil’s 5,800 counties, i.e., 58%. 29 See Lavinas, L., Programas de Garantia de Renda Mínima: experiências brasileiras, op.cit. 30 Different technical surveys agree in identifying 5 million families as the potential target

population for a national guaranteed minimum income programme. Estimates by this author are found in “Combining Compensatory and Redistributive Benefits: the challenge of social policies in Brazil”, Rio de Janeiro: ANPOCS, mimeo, 1999, page 37. In fact, this entails the narrower universe of extreme poverty (or destitution), i.e., families whose income deficit prevents them from acquiring the minimum required daily calorie intake.

9

reducing poverty, enhancing attitudes to schooling and improving the scholastic performance of scholarship holders.31

Also on the initiative of the Federal Government, the bolsa-escola has become an

important instrument in combating the use of child labour in coal mining, sugar cane plantations, sisal harvesting and other activities, within the ambit of the PETI (Program for Eradicating Child Labour), helping to replace the small but indispensable income contribution that the children provide through their arduous and high risk work. This amount was equivalent to R$50 (USD$27) until April 2000 when this benefit was cut in half by the government arguing that it was excessive for the rural areas.

More serious, however, are the overtones of patronage that has crept into the cash transfer programmes in some cities, jeopardizing a form of combating poverty whose positive impact has proved effective so far as in Brasília. This is the case in Rio de Janeiro, where the government is developing, simultaneously, two types of minimum income programmes in the state. One of them, of an experimental nature, involves a small number of children (a thousand) in two shantytowns in the city. It is coordinated by the State Secretary of Education. Parallel to this, the state government, going against century-old principles of citizenship, is distributing among evangelical churches an open “citizen’s check”, which functions like a food stamp. The check is worth a hundred reals (U$60.00) which can be exchanged for food or other goods in shops registered in the network of government suppliers. The evangelical churches in turn select among their faithful those that will benefit from this “gift” from the state, based on supposedly religious, moral and behavioral criteria, completely disfiguring a programme which has already demonstrated its efficiency and effectiveness in combating poverty, and adding a new twist to social policies in Brazil.

In view of the high degree of distortion to a social programme in Brazil whose innovative and effective approach has proven to be very successful when applied judiciously, this is an opportune moment to carry out a new and thorough appraisal of bolsa-escola programmes, highlighting their pros and cons and their potential vis-à-vis the inadequacy of traditional forms of social policy.

2.1. The Brazilian Bolsa-Escola scheme

To provide the basic features that characterize the Brazilian bolsa-escola schemes,

we will summarize the main aspects of the largest and most well known programmes, like the one of Brasilia (Federal District), Belo Horizonte (the state capital of Minas Gerais) and Recife (the state capital of Pernambuco).

31 To measure the impact of this programme, if any, the the federal government has decided to

carry out in 2001 an evaluation of the results of the program, in order to face budgetary and operational difficulties in extending its coverage nationwide, which is scheduled for 2003.

10

(a) Target group selection: Criteria and conditions Among the characteristics of these programmes are the eligibility criteria and

conditions. There are three basic criteria and conditions for enrolling in the programme: a per capita family income below a certain amount, equal to a maximum fixed local income (the most commonly used poverty lines vary from one-third to one-half the monthly per capita minimum wage, or US$25-37); school-age children regularly enrolled in primary school; and residence in the respective municipality for a minimum period varying from two to five years.

Given that such families have unstable incomes, with wide seasonal fluctuations

(they frequently work in the informal sector or outside the labor market), proving the family’s income level is always difficult. Thus, a scoring system was developed to prevent fraud and ensure proper selection. In addition to income, other restrictive criteria are adopted, including such aspects as the applicant’s living standards, household assets, possession of consumer goods, and level of schooling, to name a few. The process does involve home visits to selected families to estimate their true degree of poverty. Such eligibility criteria create large “missed population groups” that may often be just as poor.

There is a common denominator in the conditions: they all involve regular attendance by children in the public primary school system. Compulsory monthly attendance is usually 90%. Families of children who miss over 10% of classes theoretically lose their right to the stipend. In the Federal District, failing and repeating the year also means losing the equivalent of one minimum wage deposited annually in an escrow savings account as an award for adequate school performance (passing). In both cases, it is inferred that the family must guarantee that children meet the school attendance requirement and pass to the following grade. This appears reasonable with regard to the enrollment requirement, but there are doubts as to the relevance of passing, since success in school involves a series of factors outside the family scope, such as quality of teaching, the learning method, teacher training, and proper assessment of school performance by qualified specialists. Such difficulties explain why officials running the programmes are not very strict with regard to the latter criterion.

Some local governments also impose conditions on the adults, like participating at meetings organized by the programme’s administration and enrolling in job and income generation programmes offered by the local or state agency responsible for paying the stipend. These conditions may involve merely registering the unemployed family members in the National Employment System (Sine). The goal is to encourage personal development by participating family members to foster their ability to break the vicious circle of poverty and to seek reassimilation in the job market, even if only in informal or seasonal work.

Except for school attendance, which is easier to monitor through the children’s report cards, few such conditions are monitored, since communication is limited between the programme’s administration and other agencies. An example is registration with the National Employment System (Sine). Fluctuations in the labour market, with a sharp rise in unemployment, would often obfuscate such a condition where beneficiary families are

11

concerned given that their extremely deficient schooling (or even total lack thereof), lack of professional experience and skills, and social isolation render them even less employable than most. In fact, registering with the Employment System serves much more as a broader social inclusion mechanism (awareness of a specific public service, accessibility criteria, etc.) than as a tool for occupational inclusion (an aspect which is positive in itself but which fails to address the issue of poor people’s limited employability).

(b) Value of the benefit and length of time in programme Income in such programmes32 is transferred in two ways: by stipulating a fixed

sum in each case, like granting a minimum wage to each participating family, or by progressively complementing per capita family income until it reaches a predefined level, in accordance with specific needs. Except for the 45 counties in the state of São Paulo, most of whom adopted benefits of variable amounts in order to supplement per capita family income, the other programmes transferred a minimum wage (R$136, or US$75), or half of it, which reduced the programme’s fixed costs and facilitated its implementation.

Another common feature is the formal limit of a maximum length of time of the

families as beneficiaries, regardless of their socioeconomic situation, when the deadline expires. Families are usually allowed to stay in the programme for one to two years. In practice, tolerance prevails, that is, the terms are waived. This binding of the benefit to the term “school year” thus denotes the programme’s essential idea of ensuring attendance by poor children in the public school system. In the case of Brasilia, despite the fact that initially the maximum stay in the programme (set by law) is 12 months, with a possible extension for another year, more recent legislation33 has altered this term. Each family now joins the programme for at least two years and may also stay on for the period in which the child is still enrolled in primary school. Meanwhile, in Belém, capital of the state of Pará, the cash transfer is only guaranteed for one year, leading to a high turnover in the receipient families reducing the programme’s social impact in combating poverty, and affecting the goal of a 100% primary school enrollment rate.

Although most programmes have adopted the scholarship or school stipend

approach, none of them has actually guaranteed maintenance of the full benefit throughout primary school attendance (seven years). This exposes a basic weakness in the programmes in respect to their explicit long term objectives along with a shaky commitment to compulsory school attendance on the part of the government.

Finally, the only criterion leading to automatic exclusion from the programme is

the fraudulent declaration of a lower income. However, the rate of cheating has been

32 The vast majority of Brazilian programmes have adopted cash transfers as a form of benefit,

although there are rare exception, like distribution of basic foodbaskets. Such is the case for the state of Goiás, where 110 thousand families receive a monthly quota of foodstuffs, the monetary value of which is minimal.

33 Decree no. 19.391, July 3, 1998.

12

relatively low (less than 3% in both the Federal District and Belo Horizonte, the two largest programmes).

(c) Administration co-ordination and follow-up In general, technical teams have been chosen to head the programme within

specific existing agencies (e.g., the Departments of Education or Social Action). Although some authors favor the creation of an executive group directly linked to the respective executive branch (in this case the Mayor or Governor) that would act independently and quickly by centralizing the programme in conjunction with other social policies, which has not been borne out by the experiments analyzed in this study.

To date, such programmes have been run by small teams consisting mainly, if not

exclusively, of social workers, predominantly women. In general, adopting such programmes does not involve an increase in budget expenses (namely staff), since no new employees have been hired: employees have merely been shifted from one activity to another (a practice found in all cities under study).34

As for coordination with other social programmes and policies, in Belo Horizonte

(capital of the state of Minas Gerais) the scholarship target group was included in other emergency programmes, such as prevention of child malnutrition, community foodbasket (access to outlets selling food below market prices), literacy programmes for teenagers and adults, subsidized home-building programmes, and some vocational training courses. This point is highly relevant, since we are aware that social groups living in destitution fall outside the broader social protection network and are rarely or never reached by most relief programmes. The Belo Horizonte City Department of Education, which runs the programme, has thus led other city agencies (like the Food Supply, Health, and Urban Planning Departments) to adapt their programme and give priority to the demands raised by the poverty status of the population receiving school stipends.

However, little is known about the real impact of the scholarship programmes on

the children’s school performance, the amelioration of the families’ poverty, or their integrating effect as a mechanism against social exclusion. To date, only one programme is undergoing an indepth evaluation process, namely in Recife35 (1,600 households).

(d) Scale and budget participation The greatest challenge for local minimum income programmes has been to reach

a scale that will truly allow them to serve the entire potential target population. Given the budget restrictions and high poverty rates, many local governments reduce the scope of the programme and thus its coverage. This in turn also restricts its efficacy as a tool to combat poverty, and gives rise to new forms of exclusion among those segments of population already excluded. In addition to establishing restrictive selection criteria, the most frequent mechanism used to legitimise insufficient coverage is to restrict the budget allocation for funding the programme to 1%, or at most 2%, of the municipal revenue

34 Temporary employment of interns is only used for home visits and preparing the register of

families. 35 Research project co-funded by ILO-Brazil and The World Bank, coordinated by IFP/SES staff.

13

prevailing when the legislation is passed. Meeting the entire potential demand would mean committing extremely high shares of the tax base, making it impossible to implement the school stipend programme. This underscores what are often insurmountable limits on local initiatives.

According to a previous survey36, the programmes currently under way in the

form of cash transfers reach some 80,000 families, an almost negligible figure considering the total number of poor families in Brazil (ranging from five to 14 million, depending on the study referred to). Except for the program in the Federal District (bolsa-escola), which succeeded in combining a significant stipend (one monthly minimum wage) with a coverage of 70% (since the number of families living below the poverty line in Brasília is relatively small, 35,000, and the current tax revenue is relatively high as the city is the national capital) other city governments have implemented what are basically quasi-experimental or pilot projects in the hope of eventually turning them into real social programmes.

By way of example (Table 1) we estimated the number of poor families in two

municipalities where the scholarship programme has been implemented and performed simulations on the planned targetting and degree of coverage achieved. Belo Horizonte, capital of the state of Minas Gerais, is located in the most highly developed area of Brazil, while Recife is the capital of the state of Pernambuco and the largest metropolis in the Northeast, one of the regions with the greatest concentration of poverty in the country.

Table 1 TARGETING SIMULATION

total % total %

Intended beneficiary families (poverty line< R$60,00) 99 939 100% 59 319 100%

Intended beneficiary families with children in the 7-14 age group (poverty line< R$60,00)46 016 46% 28 293 48%

Potential target families 8 748 9% 15 884 27%

Beneficiery families (coverage) 1 604 2% 4 598 8%

Source: Pnad/1997 and Programa Bolsa Escola de Recife e Belo Horizonte.

Prepared by Lavinas, L. and staff, IPEA.

Recife Belo Horizonte

We note that coverage in both cities is totally insufficient (2% in Recife and 8% in Belo Horizonte), given the high poverty rates. Even more worrisome is that the criteria for over-targetting (time of residence, etc.) end up excessively restricting the programmes’ target population. Over half of the families are excluded because they have no school-age children, penalizing small children and the elderly, and over a third of the potential target population in Recife and over a fourth in Belo Horizonte are excluded again because they fail to meet the local criteria for enrolling in the scholarship programme.

36 See Lavinas L. and Bittar M., “Obtaining........”, op. cit

14

(e) Impacts – preliminary evidence in the Brasilia Bolsa-escola From the point of view of school attendance and repeat student rates (the most

important factor in drop-out rates in Brazil), there is no question that the program has been successful. Preliminary assessments indicate a zero drop-out rate among students receiving stipends in the Federal District, while the rate is 7.4% for those not on stipends37. The repeat student rate was less than 3%, as compared to 18% for children not on scholarships38.

What are the impacts in terms of reducing the vulnerability of participating

families?

There are two ways to estimate poverty: by measuring its scope (number of poor people) and its degree (the gap between the poor and the non-poor). Graph 1 shows the impact in terms of reducing the degree of poverty as a result of a monthly transfer of one minimum wage, as in Brasilia.

Graph 1 is divided by the line corresponding to the reference value used in

Brasilia to define the target family, that is, a per capita family income of less than R$65. The yellow curve distributes all participating families (24,410) in deciles, placing them in order by per capita family income before receiving the stipend. The green curve adds the R$130 stipend to the per capita family income. So, Graph 1 highlights the income gap compared to the reference value adopted in Brasilia.

One notes immediately that there is a broad range in the degree of poverty. Some

households have incomes close to zero, while others have a per capita monthly income of almost half the minimum wage. This illustrates the heterogeneity of the poor population in Brasilia, obviously making it difficult to establish a single average value for the stipend with a high degree of optimization.

A second finding is that the school stipend programme in Brasília appears to be

properly targeted, since 93% of the participating families are actually below the reference value considered as a living minimum. As Graph 1 shows, 60% of the 24,400 participating households have a per capita family income of under R$ 40.00 a month. As an example, a per capita consumption of 2,000 calories a day in the Federal District, based on the basic foodbasket defined by Federal Act no. 399/38, costs an average of R$59 a month in the first six months of 1998. Therefore, over three-fourths of these households lack the funds to meet family members’ basic food security or calorie requirements.

Only 7% of participating families have funds over the upper limit. These are

households in the last decile of the distribution (the upper tail of the yellow curve). This

37 Caccia Bava, S. “Renda Mínima – alcances e limites de programas municipais de renda

mínima”, in Programas Municipais de Garantia de Renda Mínima no Basil: impactos e potencialidades. São Paulo: Polis. Pólis.

38 With respect to the effects of the Bolsa-escola on shooling performance, no assessment has been made in Brasilia in order to state how far it goes.

15

tolerance in selecting households with incomes over the cutoff line is a conscious decision by the government in Brasilia, which did not condition the selection exclusively on the basis of income level, but weighed other factors in demarcating the target group, including housing conditions, access to public services, employability of adults, etc. It may thus be inferred that the degree of vertical spending inefficiency with the guaranteed minimum income programme is relatively low in Brasilia, given the programme’s scale. This is because the percentage of participating households that do not belong to the priority group is small (considering only the income criterion, which we know is insufficient to accurately define poverty).

On the other hand, there appears to be no signs of horizontal inefficiency, if we

consider only the target group using the eligibility criteria: the number of households that qualified for the program (35,000) is quite close to the estimated potential number based on secondary data (PNAD). In fact, the problem of horizontal inefficiency arises with households with insufficient income that do not meet all the pre-requisites of selection particularly that of more than five years’ residence – and thus fail to qualify for a school stipend. While there are isolated cases of local initiative (outside the national programme benchmarks), it will be impossible to avoid adopting mechanisms that cause horizontal inequality. This is a means by which local and state jurisdictions may avoid problems of “importing poverty”,39 but they cannot avoid other problems such as generating inequality among the poor, which is an unfair outcome that must be addressed.

The green curve shows a significant improvement in family income once the

school stipend is added. The number of households with per capita family income below the reference value, drops from 93% to 50%, in other words, by half. In absolute figures, this means that more than 10,000 participating households are above the established cutoff line in the Federal District. In relation to the degree of poverty, the graph shows how the income gap is distributed in percentage terms, compared to the reference value adopted by the programme.

39 See Lavinas and Varsano, “Minimum Income Guarantee Programmes and Coordinated Action

Against Poverty”, text for debate n.534, IPEA-DIPES, 1997.

16

Graph 2Income Gap of Households Benefiting from Brasilia MIGP in relation to Half a Minimum Wage Value

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

100%

120%

140%

160%

180%

200%

220%

240%

Tenths

Ga

p

YFPC

YFPC with R$130,00 benefit

Sources: Brasilia MIGP register (September 1998 version)Notes: 1. YFPC - Household Income Per Capita 2. Number of beneficiary households: 24410

The income gap of the poorest 10% narrows by 30%. As the green curve shows, this improves over the distribution, due to the smaller size of the household group. The narrowing of the poverty gap is a surprise: the fifth decile of the distribution, 50% away from the reference value (half a minimum wage), before including the monthly stipend to the household income, practically eliminates this gap. On average, the poorest 50% narrows by one-third the income gap between them and the one-half minimum wage per capita value. The distortion is due to the last decile in the distribution, with net income of over 80%, indicating that some households have succeeded in escaping from poverty, at least temporarily. With the guarantee of more income to poor families, the guaranteed minimum income programme attenuates insufficient monetary income, one of the main dimensions of poverty.

(f) More efficient social spending and more effective social policies The most interesting and irrefutable aspect of the school stipend programme in

Brasilia lies in the increased efficiency of social spending in general, i.e., not only of the portion allocated to the guaranteed minimum income programme. We know that the poorest population is rarely included in the vast majority of social programmes and is also excluded from public policies in general, as if the poor did not exist or their demands were ignored. Precisely for this reason, another structural dimension of poverty lies in the lack of access to goods and services by the most disadvantaged social groups.

One undeniable feature of the guaranteed minimum income programme is that of

increasing the potential to the redistributive impact of social policy as a whole, acting in an integrated and integrating manner on the poor population’s living conditions. Proof of

Graph 2 Income Gap Households Benefiting from Brasilia MIGP

in relation to Half a Minimum Wage Value

17

this is the table based on a methodology40 created specifically for this purpose, which takes into consideration the nature of social spending by sub-program and sub-activity.

As Table 2 shows, social spending on actions focused on eradicating poverty (D)

has increased in the Federal District faster than social spending with an impact on poverty in general. The ratio (D/C) of the former over the latter increased significantly in just three years, from 3.80% to 11.7%. As a result, there was a sharp rise in per capita social spending on anti-poverty measures (from R$104 to R$279), while the same was not true for general per capita social spending.

This was possible because when households are registered in Brasilia’s

guaranteed minimum income programme for monthly benefits, they become a priority target for other state government social programmes. They receive preferential treatment in outpatient clinics, health care centers, and schools. They acquire faces and names, so to speak, and earn the status of citizens, albeit poor ones.

The scholarship programme fosters coordination of different policies, induces

synergism (previously absent) between government agencies, cuts through red tape in public administration, and places the poorest and unassisted population at the center of social action. Thus, the entire process also becomes more effective.

This may be the most relevant impact of the school stipend programme: reorienting

social spending towards the most excluded population groups, promoting mechanisms to provide universal basic public services.

Table 2

PER CAPITA SOCIAL SPENDING IN THE CAPITAL OF BRAZIL, BRASÍLIA* 1995** 1996 1997

Resident Population (A) 1,706,000 1,821,946 1,865,352 Poor Population (B) 279,000 297,962 360,946 (B)/(A) 16.35% 16.35% 19.35% Total Value of Actions with Impact on Poverty (R$’000) C) R$ 68.204,98 R$ 271,035,88 R$ 861.017,69 Total Focused Actions Against Poverty (R$’000) (D) R$ 29.229,59 R$ 71,763,58 R$ 100.741,04 (D)/(C) 3,80% 9.56% 11.70% Per Capita Social Spending (R$) R$ 450,30 R$ 412,22 R$ 461,58 Per Capita Social Spending on Anti-Poverty Activities (R$) R$ 104,77 R$ 240,41 R$ 279,10

Source: SIAFEM: Department of Finance and Planning, Federal District (Brasília) and Rocha, S. Prepared by: IPEA/DIPES, Lavinas, L., 1998 * Includes staff (payroll) expenditures ** Values in R$ as of 1996 according to the annual average IGP-DI.

2.2. Summary of the Brazilian case

The study of the Brazilian case shows that minimum income guarantee programmes to alleviate poverty have been implemented, but only in a very fragmented and difficient way. Actions have been excessively selective and reduced in scope, timid

40 This methodology was used in the Annual Balance Sheets for the Budget Implementation

prepared by the Department of Finance under the government of the Federal District.

18

even, and lack continuity and expansion, mostly due to the limitations of funds and the budget restrictions of the inflation stabilization plan. Their implementation has depended on the initiatives of progressive politicians and social scientists, rather than a widespread demand of the population for their citizenship rights or from ethical considerations. Moreover, their objectives have been rather short-term since they lack a strategy to insure the universalization of basic education, or a definite erradication of poverty. Due to this myopic approach they also impose limits to the permanence of families in the programme and require several additional commitments from participants, which end up increasing the risk of failure. In short, they still have what Lo Vuolo called target- assistencialist bias. In spite of all these shortcomings, they are still very appealing because of several positive aspects:

First, these programmes give visibility to the excluded – granted, up to now, only

to a small fraction of them – making evident through the selection mechanism they employ, the profound degree of exclusion and the extreme poverty and insecurity to which tens of millions of Brazilian are subjected.

These programmes also innovate by introducing a redistributive component in

compensatory policies: in addition to cash transfers, they improve the welfare of the population that lives outside the public services system by increasing the level, improving the allocation, and widening the scope of social expenditure.41 This happens because they induce new demands for social services from those that never had access to that system, and this deficient coverage is both widespread and acute in the Latin American poverty scene. The demands of the beneficiaries of the school scholarship programme induce the increase of targed social expenditure and strengthens the redistributive dimension of the programme. Brazil is a very unequal country, where the richest 10% of the population appropriate 47.8% of the income, and where only 7.5% of the active population (5.5 million individuals) has a college education. In countries like this, programmes which combine compensatory and redistributive effects are important mechanisms for spreading access to citizenship and for the insertion of the excluded.

Finally, these programmes offer a more dignified alternative to workfare, which

has been throughout the Twentieth century almost the only way to guarantee the survival of the poor and rural workers, who were thereby transformed into second class workers with wages that were only a fraction of the minimum wage. They are also an alternative to the clientelist distribution of food baskets which up to now constitutes the basis of the Brazilian social compensatory policy. They are, therefore, a good substitute to the old assistencialist policies and could, preferably, be called redistributive target policies.

The preliminary evaluation of some of the programmes studied here shows that

the activity rate of the adult population receiving the benefit remains stable, while the unemployment rate is reduced, which implies that the number of employed workers has increased. Minimum income policies also reduce economic insecurity because they provide a safety net to those that cannot claim unemployment benefits. In the less favored

41 O gasto social eleva-se a 19% do PIB brasileiro, muito embora 2% apenas sejam alocados em

políticas e programas de combate à pobreza.

19

social groups informality and unemployement rates are much higher than the national average.

Perhaps this conjuncture has unwittingly given us a unique opportunity to devise,

in a situation of heavy budget restrictions, continuous structural poverty and rising unemployment, alternative solutions to the serious and persistent social policy problems. There is no doubt that MIPs can contribute to this effort, radically changing the profile of Brazilian social policies against poverty. They are not a solution in themselves, but are a good start to change precisely the feature of these policies that for so long have been unable to produce significant changes in the lives of the poorest: the lack of accessibility to public goods and services, which affects so deeply the welfare of the needier population. In this way, the MIPs may induce the redistributive impact of social policy as a whole, integrating them in the life of the poorer and underprivileged classes of the population.

It is nevertheless essential to change their design, by eliminating all excessively

selective criteria that generate horizontal iniquities among the poor, and ensuring coverage to all the potential beneficiaries. Therefore elegibility should be unconditional and the scope nationwide.

3. Minimum income and its spread in Latin America

The appeal of minimum income programmes has grown considerably in Latin America in recent years prompted by some preliminary evidence that this form of social action has positive repercussions by reducing poverty in the short term and upgrading education among social groups excluded from the so-called social protection system. The following is an overview of the Latin American debate and of the design of existing programmes, like Progresa in Mexico, the broad scale of which merits special attention.

3.1. The Argentine debate: new rights for childhood or focusing on poverty?

Argentina debated a bill to create a Citizen Income Fund for Childhood (Fincini)

in the late 1990s. Its objective was to finance a monthly single value bonus (Incini or Ingreso Ciudadano de la Niñez), for all Argentine girls and boys under 18 years. The project does not entail any type of condition to obtain such a benefit, on the children or families. In other words, this is a global benefit, without consideration to selection or income level considerations, nor does it entail obligations of school attendance, etc.42

42 In Argentina there is a family assignment programme (Asignaciones Familiares), consisting of

the monthly payment of a benefit ($50) to wage earners, plus an extra annual payment for each child attending school. The problem is that this subsidy covers only approximately 45% of children, those whose parents are in the formal labor market. The dependents of informal workers or unregistered wage earners, etc., are thereby excluded. To do so, the incorporation of new budget resources was projected at first estimate at 1% of the GDP.

20

At the same time actions taken on the fight against poverty are decided at a provincial level, as in the case of income transfer programmes for needy families. In Mendoza, for instance, a pilot programme was implemented in 1998, initially affecting 5,000 from the 44,000 families that are already somehow benefiting from other social programs and, for this reason, have been registered43 by the Social Development Secretariat. This is the estimated number of poor families in the province of Mendoza. Initially, only five of the 18 municipalities that form the province were considered for the pilot experiment. The project should help to restructure a group of activities undertaken by the Secretariat to support the population at risk. The purpose is to integrate the different social services44 in a single benefit.

In Argentina, as in Brazil and many other Latin-American countries, there are numerous policies against poverty.45 But, taken as a whole, the various initiatives are still fragmentary at best, and renewed efforts and resources in programmes of greater leverage and impact are required. For the political party in opposition, these programmes have been unsuccessful and there are calls for a broader programme that guarantees a basic income for all, a programme estimated at a cost of US$3 billion/year.

The Social Development Secretariat46 of the federal government acknowledges the relevance of a minimum income programme, although it does not agree with the universalistic application but would rather focus on poverty in the country per se. The Secretariat claims that 19% of urban households live below the poverty line, that is, they have a per capita family income of below US$155 a month.47 This would be the population that would benefit from the programme. To date, it is more probable that the sum of US$120 or $150 a month would be fixed as the benefit value for direct transfer to families, a sum that would be tied to obligations regarding infant and secondary school attendance and basic childhood nutritional requirements. It should be noted that while compulsory school is required by law, the school drop-out rate is very high among poorer families.

Fincini, on the other hand, stipulates the monthly grant of a monetary benefit, whose initial value would be fixed at $60 per child. This figure would be updated every six months, according to variation in the Consumer Price Index. Every pregnant woman would have the right to receive Incini from the fourth month of pregnancy. In the case of minors with disabilities, the benefit should double ($120), and be guaranteed for life. This benefit, along these proposed lines, would affect one-third of the Argentine population (population under 18-years old), or approximately 12.5 mi llion people. The estimated financial cost in pesos varies between 7.5 billion (on a monthly transfer sum of $50) and

43 In fact, the Mendoza Social Development Secretariat has a very detailed database on the status of families in at least one of the locally run social activities.

44 Today, several programs somehow pass monetary aid on to the poor population, namely through municipal employment programs whose preferential target public are poor families. Some offer 200 pesos a month for up to six-month contracts at the mo st in civil construction or community activities. Programs exist for distributing dried foods. Mendoza province estimates its annual social spending at around 3.5 million pesos.

45 In 1997 the federal social expenses totaled two billion dollars. There is no official data on social spending by provinces or municipalities. It is worth mentioning that 80% of social expenditure is decentralized in Argentina (only 20% is the responsibility of the federal government).

46 This secretariat is directly linked to the Presidency of the Republic and its Secretary has ministerial status. 47 1US$ = 1 Austral. Data supplied in July 1998 by the Social Development Secretariat of the Presidency of Argentina

Nation.

21

10.5 billion (alternative of $70 per month). The revenue to ensure such benefits would be obtained by suppressing and substituting some social programmes of an assistencial nature, redirecting their resources to Incini. Such a bill has still to be voted in the Argentinian Congress.

While Fincini has yet a long way to go and a debate on its effectiveness has been postponed for many years, a new initiative is gaining ground that will redefine the debate. From the first, Fincini was an ambitious idea to create a social protection network for all non-adult citizens in the form of a payment of a citizen income. The priority given to the younger, future generation was a concession made largely with a view to tax limitations and considerations, but it may nevertheless be seen as instituting a true mechanism against poverty. New limitations, however, are being proposed, this time for a programme that will address the poorer population as a whole. The Federal Government Social Development Secretariat, citing factors such as unemployment and under-employment figures among others, believe it more opportune and suitable to ensure a right, albeit of a preventive nature, for all families considered ‘poor’ (defined by the UBR -Unsatifactory Basic Requirements). We now have a new dispute, which, on the one hand, addresses a universal global childhood benefit and, on the other, a variant, giving priority to the poorer population at large.

3.2. Progresa: the Mexican three-dimensional strategy Mexico has also adopted measures to combat extreme poverty that are similar to a

minimum income programme. The benefits paid to the families are of two types: educational stipends and food supplement support, and both require that the beneficiaries fulfill several conditions to participate in the program.

The most important programme is Progresa,48 the national programme of education, health and food, that aims to improve the living standards of 4.7 million poor, predominantly rural families (28% of the population). Progresa aims to formulate “an integral response to reverse the privations of the population living in a situation of abject poverty”,49 The condition required for the grant to be maintained is that grantees attend at least 85% of the classes given during the school year. It started in 1998, and, to date, has reached 2.6 million families in 2,100 municipalities,50 in the Mexican states. The support is in the form of monthly payments to each child attending the third elementary year to the third secondary school year. This monetary aid for each child is supplemented by a set of actions to support the family in the health and food area and by granting another monthly benefit to the family to reinforce food accessibility.

One of the interesting features of this programme is that the value of the scholarship paid to each student is raised as he advances in his education, gradually rising

48 Progresa, Programa de Educación, Salud y Alimentación, 67 pages, federal government, 1998. 49 Op. cit. p.40. Item 3.2. 50 Progresa. Evaluación de Resultados del Programa de Educación, Salud y Alimentación.

Primeros Avances, noviembre 1999. Secretaria del Desarrollo Social, Mexico, p.397.

22

from a monthly value of 8.5 dollars in the first year of elementary school51 to32.8 dollars in the last year of secondary school.52 The purpose of this is to reduce the risk of the students dropping out of school as they progress through the several grades and as other activities compete with education for their time. That value is also increased by 20%-25% for girls enrolled in secondary school to compensate for the higher female school drop-out rate, at that stage of schooling due to the pressure arising from the social and gender division of work. By offering this positive differential Progresa tries to encourage girls to proceed beyond primary school in their education, rather than getting involved in domestic work. In Mexico, unlike Brazil, the school performance of girls is lower than that of boys, as a result of sexist discrimination in the family.

The Mexican government estimates that this scholarship raises the household income to more than 15% above the income level that would be reached with child labour.

In addition to this monetary benefit, grantees also receive school material and/or resources to acquire them. A health plan is also part of the programme, aimed at avoiding child malnutrition and promoting preventive health care for children at school age through scheduled health clinic visits for each family. Families with undernourished children from four months to two years of age have preferential treatment in receiving food supplements, as do women with difficult pregnancies. To fight nutrition and encourage a more diversified diet, all poor families in the Progresa programme are also provided with a monthly monetary food grant of US$13.14.

The total monthly benefit awarded to each family is however limited to a maximum value of US$80 dollars. Poor families without children or with children outside the selected age group only receive the monthly food grant. The overall average value of the monthly stipend for each family is US$27 dollars. In families that have children of school age, the average grant is US$40 dollars,53 which corresponds to about 41% of the current minimum wage in Mexico.

Initially, the length of time during which families may remain in Progresa is three years, but they may apply to remain for an additional three years, which extension depends on a thorough socioeconomic assessment.

The federal government, in implementing and financing this programme, depends on the close cooperation of state and municipal governments to develop other selected programmes that help alleviate poverty and improve the social protection network, such as community kitchens, school lunches, family planning, etc. States and municipalities are responsible for consolidating the basic social infrastructure of their regions, offering

51 Exchange of the average Mexican peso in January 2000 was 9.30 pesos/dollar. 52 Na verdade, esse valor era inferior no primeiro semestre de 1998, respectivamente 65 pesos

(US$ 7.6) e 225 pesos (US$ 26.3), sendo o montante acima mencionado o valor corrigido para o segundo semestre de 1999. Os dois extremos correspondem ao valor mínimo e ao valor máximo de uma bolsa de estudos paga a crianças cursando o ensino primário e adolescentes matriculados no secundário.

53 A título de comparação, vale lembrar que o salário mínimo vigente no México em dezembro de 2000 é de 900 pesos ou de US $ 96.7.

23

public services the demand for which increases with the fulfillment of requirements imposed on the families and students to remain in the programme. Therefore, the decentralized local expenditure in poverty reduction programmes also increase with Progresa.

The total cost of Progresa in 1999 was US$827 million,54 of which 38% was spent in payment of scholarships, 53% on food grants, and only 8% on health-related activities. Its operational cost was estimated at about 4.3% of its budget, which is commendably low, given its broad scope: it reaches three out of four poor families in the rural areas.

The evaluation of the first phase of the programme, for 1998-1999 indicated it has attained satisfactory results: a) On health: the yearly number of medical examinations per family increased from five

to 8.6, the actions in favour of undernourished children increased 30%, and neonatal care increased 16%. Coverage of health services also increased in areas of extreme poverty.

b) On education: school enrollment increased, especially in secondary education (24%), there was a reduction in child labour, especially in the age group of 12-13 years, and a reduction in school truancy.55

c) On nutrition: the expenditure on food in the assisted families was 7% higher than in other non-assisted poor families, which indicated an increase in food security, and expenditure with non-food items increased by 5% (on clothing for children, mainly).56

d) Last, but not least, its implementation did not reduce the activity level of adults, showing that the programme does not entail a disincentive to work, an effect that some theoretical analysts suggest could happen.

3.3. Ecuador: Todos los Ninos y las Ninas a la Escuela

In 1999 Ecuador was considering, within the framework of its negotiations with

the IMF, the adoption of a social programme to compensate for some of the negative social impacts of the foreign exchange and currency crisis, and the ensuing economic stabilization programme, on the poorer population.57 In Ecuador absolute and relative

54 Segundo fontes oficiais, 80% dos recursos alocados no programa constituem-se de transferências

monetárias às famílias (ou US$ 640 milhões), os 20% restantes sendo transferências in natura (combate à desnutrição infantil e suplementação materno-infantil).

55 ¼ das crianças que antes trabalhavam e estudavam, passaram a estudar apenas ao se tornarem bolsistas.

56 Preliminary Evidence of the Impact of Progresa on Consumption, IFPRI, Hoddinott J. and Skoufas E., January 2000, Executive Summary, pages IV and V, and data from a study of the International Food Policy Research Institute.

57 Segundo o documento do Ministério da Educação, Cultura e Desportos do governo equatoriano, “Perfil del Programa Todos los Niños y Niñas a la Escuela – Beca Escolar” (versión preliminar, agosto de 1999, 15 paginas), o PIB do país teve crescimento negativo de 7% em 1999, a pobreza medida pela renda per capita diária inferior a 2 US$ atingiu 43,8% da população no mesmo ano, contra 37,9% em 1998, a taxa de desemprego aberto chegou a 15% em maio de 1999 e o salário mínimo real caiu 22% entre janeiro de 1998 e junho de 1999. Op. cit. página 3.

24

levels of poverty increased dramatically during the 1990s, making the adoption of these measures imperative.

Among the several medium- and long-term policies which were considered, the

one called Programa Todos los Niños e Niñas a la Escuela, stands out because it is inspired by the successful programmes which were implemented in Brasilia (Brazil) and in Mexico (previously described).In the same spirit, it provides a school grant to families in the lower quantil of the income distribution which also have children of age 6-15, under the condition that they mantain a school assiduity rate of at least 90%. The targeted population – families with a monthly income lower than US$ 60 – numbered 286,000 families (approximately 572,000 children), of which 74% live in rural areas, where poverty is more pronounced and schooling is very far from being universal.

The three main objectives of the programme are centered in the educational

sector: to increase enrollment and assiduity in elementary school, to reduce drop-out rates, and to stimulate the completion of elementary schooling. The reasoning behind this design is by now well established: only the improvement of the size and the quality of the human capital can break the vicious circle of poverty. Education then becomes also an instrument of social mobility, as child labour is eliminated. The programme also indirectly attempted to reduce migration from rural to urban areas, which tends to increase in times of recession.

The value of the monthly stipend during the school year (9 months) is US$10 per

child,58 to be paid every two months to the mothers, through the banking system. To reduce repetition of the school year, and the consequent high levels of drop-out, an extra benefit of US$10 will be deposited in a savings account for each year of schooling which is completed by each child. Half of these savings can be redeemed after the sixth grade, while the other half is available after the completion of the tenth grade. In case of evasion or two consecutive repetitions of the same grade, the savings deposit is forfeited.

A third benefit is also envisaged which will act as an incentive to elementary

school teachers, whose absenteeism is noticeable particularly in the rural areas. It is a kind of productivity bonus which will be paid to about 16 thousand teachers monthly. Its value will be US$25. The educator contributes to the success of the programme by stimulating the student’s regular attendance and never missing class.

The annual cost of the programme was estimated to be US$38 million, whose breakdown is: US$25.7 million for scholarships, US$3.5 million for teachers, and US$5.7 million for the savings deposits. Most of this amount, however, will be financed by the World Bank and The Inter-American Development Bank. Little use will be made of ordinary budget allocations, which will be responsible for less than 10% of the total cost.

It will be executed by the Ministry of Education and implemented in a decentralized way, with the help of local school boards organized at the parish level,

58 Estimativas do governo avaliam em 2 o número de crianças nas famílias mais pobres em idade

escolar obrigatória.

25

which will be the geographic units where the beneficiary families will be selected. These parishes will be ranked at the national level on the basis of UBR to establish priorities for participation in the program. 4. Final comments

The minimum income programmes being discussed or implemented in some Latin American countries, and their spillover effects, has allowed us to revisit a debate which has been polarized between two extreme concepts of social policies, that can be summarized in the dichotomy of universalism versus selectivity. It is not a coincidence that in Argentina, where the social protection system has always been more effective and encompassing, there is more resistence to the proposals based on selectivity, and the debate regarding their design is more radical. On the other hand, in Mexico, Ecuador and Brazil, which are all countries with a high degree of social exclusion and where inequality has been very pronounced and constant in the last decades, targeted programmes have greater receptivity.

Targeted anti-poverty programmes, like minimum income and school scholarship, signal that it is possible to develop innovative strategies to strengthen universal public policies, even in an adverse framework, through the concentration of resources and of the priorities on the needy. These programmes combine income tranfers and the increase in the coverage of social policies. In this way they induce redistributive effects in the short and medium term, which foster equity. They are not based on the creation of a paralel public health or educational system to cater exclusively to the poorer population, thus offering a restricted level of public services. Rather, they stimulate their democratization, because they bring to the official system demands which have never been considered, that of the poor and economically insecure. In addition, if nationally widened and accessible, without restriction to any individual in need and socially vulnerable, to be precise, if established as citizen rights, these policies open up the way to setting up a system of universal allocations, non-existent in Latin America up to now. Indeed, the social policy in this region is still nowadays the most consummate expression of selectivity from the top, an “iceberg citizenship” in other words. The minimal income programmes may offer for the first time the possibility to discuss everyone’s right to benefit from a basic security threshold.

In spite of the strong disparity between the programmes in different activities under examination here, we can state, on the basis of some characteristics of their design, that they broaden the citizenship of individuals that were previously excluded from all social and public services. In this way they “ensure minimal income security and a broad orientation to socio-economic security for the disadvantaged”.59 Some premisses used in the design of these programmes should be shifted, like the over-targeting mechanisms, its local coverage (non-national) and temporal limits in the participation of the target population. Pre-conditions – apart from school attendance – should also be canceled out

59 The ILO InFocus Programme on Socio-Economic Security . A Medium-Term Workplan (SES), miméo, 24

páginas, Geneva: ILO, 1999.

26

since they are not effective and needless. They could then become a truly “innovative alternative to enhance the coverage of conventionnal social protection systems”60.

60 Idem.

APPENDIX 1

MINIMUM INCOME PROGRAMS - 1999Federal Government - Ministery of Education

States Municipalities Mean Children FamilyStipend (R$) Beneficiary Beneficiary

AC 2 43.46 4 719 2 202AL 14 39.39 34 983 14 673AM 5 48.18 8 241 4 333BA 150 36.55 262 455 126 019CE 37 34.73 56 844 29 810ES 40 43.20 53 851 23 363GO 31 33.62 13 705 6 893MA 22 39.16 46 815 21 888MG 215 39.37 207 325 104 462MS 15 37.01 4 390 2 495MT 18 43.44 14 312 6 312PA 12 51.90 15 870 5 314PB 30 36.82 39 068 19 893PE 43 34.51 96 068 47 821PI 11 36.20 5 827 2 446PR 42 29.52 11 706 6 361RJ 32 30.09 23 543 13 254RN 42 39.79 39 538 20 489RO 6 49.17 7 501 3 168RR 1 34.61 948 423RS 50 35.27 10 938 6 021SC 39 29.15 11 541 7 247SE 16 44.23 24 012 10 199SP 120 30.37 88 281 16 731TO 12 37.37 5 589 2 463

TOTAL 1 005 37.47 1 088 070 504 280Total Federal Spending R$ 3 960 146 724

Source: Ministério da Educação, 1999.

Note: 1US$ = R$ 1.80 in 1999.